Chinese Serif Italics

There is currently no Chinese equivalent of italics used for emphasis. Simply slanting the characters is not only inadequate (and bad-looking), as in Western languages, but ignorant of the history of Chinese writing, where body text was in columns and only display script was arranged horizontally, which would preclude the slanting of vertical strokes. It looks bad.

“That under heaven there is nothing uglier than fake italics is something that thou knowest, venerable sir.”

In short, what we currently have falls short of the demands of modern type, especially with the common use of Microsoft Word, because there is no satisfactory solution already being used to fill the need. This isn’t even optical sizes and other fancy options that Adobe provides with some of its Western fonts. This is italics I’m talking about. So what options are there?

Seal script display faces are too different from standard Ming, Song or Fangsong faces to be used with text faces as emphasis markers, being more similar in connotation to blackletter in Western type, except less legible to the modern reader. Xingshu (semi-cursive script) and caoshu (cursive script) are too decorative and not legible enough in small print sizes; kaishu (regular script), while calligraphic without looking overly scripty, is too neutral to serve as an italic.

Enter lishu (clerical script), now used primarily as a display face but highly legible. It usually signifies antiquity or artistic flavour, but of the candidates lishu has the most potential as a starting point for developing an italic typeface as a companion to the “roman” serif typefaces.

  1. It is legible enough and traditional enough to complement serif fonts.
  2. Its flair is comparable to that of italics in Western type, and there’s more of a sense of serif to speak of than with other calligraphic styles.
  3. Whispiness will not be a problem.

However, there are some major challenges. Clerical script has very horizontal proportions compared to regular script and serif typefaces already in standard use, and its colour would appear too dark on a page of Ming or Fangsong type. There are also some baseline issues with clerical script more prominent than with regular script.

So we need a lighter, less horizontal script than the clerical script already in existence that also doesn’t clash badly with serif fonts. Compatibility could be determined by comparison with kanji, hiragana and katakana integration in Japanese serif fonts .

If anyone with the requisite knowledge of Chinese calligraphy and general typography will do what it takes, my eyes will thank the type designer when I start seeing the new typeface replacing the ubiquitous faux italics. I’ll be a tester if you want.


15 responses to “Chinese Serif Italics

  1. Kăishū can be a candidate. There are different styles of kăishū: liŭtĭ is one of those that are elegant, legible, and do not take up much proportion horizontally.

    Let me know if you would like to discuss more.


  2. I wonder if it stands out in a way that would immediately mark it off if it were used for emphasis. I know Chinese does have emphasis dots, but almost no one uses them in type, and instead everyone, because of word processing software, hits the “italic” function.

    Since the typeface designers have furnished italic fonts anyway, they might as well do something better than just slanting characters and calling them italics. For a Song face, maybe fangsong would work as a set of italics.

    The thing I found interesting about lishu was that the curves and stuff reminded me of the essence of Western italics. Both are older calligraphic forms from about the time that printing emerged.

    But as far as I know (which is pretty little), no major calligraphic style has developed since the maturation of kaishu in the Tang Dynasty.


  3. Hello, why not communicate with actual people in Hong Kong and China and see what solutions they come up with? I am sure they have come up with solutions of their own. If you keep thinking within the western realm, you’re going to come up with a solution that Chinese themselves are not using because they already have a solution!


    • As I note above in the comments, traditional printing generally employs emphasis dots. Whenever I see people from Hong Kong and China typesetting their materials, however, they simply hit ‘italic’, even though slanted characters are entirely opposed to the nature of Chinese characters. Perhaps people have proposed solutions, but I don’t have the connexions.

      But this post is also a plea for Western firms including Chinese characters in their typefaces to recognize that slanting is an inappropriate solution and especially to stop making (and selling) font files that are essentially the characters slanted to a certain angle rather than creating a real solution. Obviously, this is a lot of work, but I believe we need it.


  4. Christopher Miller

    A very interesting question, indeed. I’ve been thinking of these things for a while now in connection with my interest in developing a sign language script. Because of the generally simultaneous structure of signs, a linear script is ill suited to the task and the best approach would be to follow the general structural principles of CJK scripts, taking Hangul as the inspiration for the method for combining constituent characters into a single block representing a sign syllable.

    Since the base characters are designed following CJK stroke structure principles, I would expect that Chinese, Japanese and Korean deaf people would likely take to the script like fish to water. I expect the learning curve would be a bit steeper though for users in non-CJK-using parts of the world.

    Since the goal is for the script to be used for any sign language, I keep in mind the importance of allowing for typographic conventions those who don’t use East Asian scripts are familiar with, including variants that correspond to italics, bold, capitals and small caps. Including these in the script would provide some measure of familiarity in a writing system otherwise very different from what western users are used to working with.

    It’s interesting that these are all developments of the European tradition that have spread from Latin script to the culturally connected Greek and Cyrillic scripts, but aren’t used, as far as I know, in any other script of the world, linear or otherwise. Capitals and italics are essentially different styles of the same script that were combined with the basic script descended from uncials, but with special functions. Small caps were developed — I assume — to avoid the overly loud appearance of capitals compared with the rest of the text.

    It seems to me that CJK typography might be open to adopting conventions analogous to italics and capitals, for similar functions; after all, the traditional vertical columns moving from right to left have been abandoned for many or most purposes in favour of European style left to right lines moving downward on the display surface, and this mostly since the 1940s.

    I notice that italics have two characteristics that set them apart from roman glyphs. First, of course, is the slant from the vertical, which angling tries to imitate. Second, the characters (especially in the most classic and best designed fonts) tend to be narrower than their roman counterparts and (at least I get this impression) are more closely spaced as well, so that a line of italics is around 80-90% narrower than its roman counterpart. This is less often the case with sans-serif font families, Gill Sans being a notable exception that acts like serif families.

    The closest equivalent to this in CJK is kai-type fonts: the characters are narrower and their horizontal strokes are drawn more relaxed, at an angle of as much as some 10˚off the horizontal (analogous to the off-vertical angles of Latin italics). They also have a tapered structure on downward strokes that again helps gives them a more relaxed appearance than the blocky Song/Ming type glyphs, much in the way the curviness of italics sets them off from roman glyphs.

    In old-style vertical text, centred in the vertical line, I think kai fonts would do very well as an analogue of italics, both because of their movement and their overall thinner appearance. In horizontal layout, they could also work provided the characters are spaced more closely in the way italics are in Latin script.

    Your suggestion of some variant of clerical characters as an analogue to italics was very interesting. Personally, I think they are a better analogue for capitals (and small caps). Like I said, kai glyphs contrast visually with Song/Ming glyphs in ways very similar to the contrast between italic and roman glyphs. The visual contrast between clerical script/lishu glyphs and Song/Ming is very analogous, in my mind, to that between capitals and minuscules. Historically, minuscules developed through simplifications to the original capitals, and although the relationship between lishu and Song/Ming glyphs is not the same, lishu is analogous to capitals in being the older script — with the additional analogy in size contrast.

    Capitals are larger than their minuscule counterparts, occupying more vertical space than the main body of most minuscule glyphs. Analogously, lishu glyphs are wider and take more horizontal space than regular Song/Ming glyphs.

    If you look at the overall distinctions in the two script traditions, they both revolve around how script types interact with the axis perpendicular to the overall writing direction. In the left to right Latin script, capitals contrast by filling vertical space more than minuscules do; italics contrast by their slight angle off the vertical (together with their narrower appearance). In CJK, traditionally written in top to bottom in columns, lishu glyphs fill horizontal space (perpendicular to the writing direction) more than regular characters do, and kai and schoolbook characters are more relaxed in their angle off the horizontal. If CJK typography were to adopt these as functional equivalents to Latin (small) capitals and italics, I think the visual clues marking them off from the main text would be very similar to those in Latin scripts (and immediately obvious to those familiar with Latin script languages).

    Now, if only the Latin typefaces included in most CJK fonts could be better designed (and include italics, to boot)! I also hope to see the day where word processors would allow vertical CJK layout to include horizontal rtl or ltr text, with any columns interrupted by such text “jumping over it” just like text already does with inline graphics.


    • Wow, thanks for a quite detailed and helpful comment.

      My concern about non-Song/Ming glyphs is their reduced legibility at smaller sizes. At least some kai faces turn into globs of jelly at sizes at which Song/Ming faces are still fairly legible.

      Best wishes on developing a sign language script!


  5. 我是香港人…
    I can’t agree more with Umm… // Saturday, 6 June 2009….in China and Hong Kong, we don’t readily use MS Word the same way you do!

    I suspect you are yellow outside, and white as white inside. That’s what America has done to you. I recommend you travel to Hong Kong for a few months and then Beijing for a few weeks.

    You are overly negative on Chinese, I suspect you are doing this on purpose.


    • 別用叚的 url 吧.

      What I do see are Chinese people here using MS Word for their documents and adopting the default Western methods without recognizing the differences between typesetting Western languages and typesetting Chinese.

      I speak Cantonese at home the same as you do – well, maybe not the same, because I haven’t acquired the sound changes that have taken place in Hong Kong Cantonese since the 1960s. Unfortunately, I’m unable to travel in Hong Kong and mainland China, although I’m still at liberty to go to Taiwan.

      As for negativity about Chinese people, I do know a lot of people hate being Chinese and carry a lot of shame about being Chinese and wish they weren’t Chinese. I do admit some days I’m embarrassed to be Chinese, as some days I’m also embarrassed to be American. I merely hope, however, that Americans and Chinese will equally be willing to examine their weaknesses honestly.

      I do object to the constant suspicion on the part of some Chinese people that whoever is Chinese and says less pleasant things about his own people is somehow less Chinese. Surely you aren’t hoping I’ll make the kind of comments about the bulk of Chinese people the way Jackie Chan did a few months ago.


  6. I really like this article. I never thought about chinese italics until only very recently and then your article comes along and reaffirm my sentiments. Italics is clearly a western influence on asian script. So is horizontal writing for some. We are so westernized that hardly anyone thinks chinese italics is strange at all anymore. I am currently looking for italicized asian script in-use through asia be it singapore, japan, korea, thailand, arabic, tamil, etc. I would like to investigate the ‘acceptability’ of italicization of respective asian script in their respective mono/bilingual/cosmopolitan culture. If you have any leads, please let me know. Thanks!


    • In Arabic, most of what we see is italic, and kufic is roman. Or perhaps Nastaliq is the most italic, since it really is used as a favoured text type for Persian and Urdu.

      For Thai, maybe this.

      The east Asian scripts, all written with the same kind of brush with the same squarish space will have no conventional italic. Also see Mongolian: the more ‘italic’ writing, again, keeps the baseline unaltered.


  7. I agree with you that the equivalent of italics in CJK is kai. Italics is not simply slanted type, but a stylized version of a calligraphic hand that is cursive but not to cursive, so kai perfectly fits the description. The problem is kai is, as you said, the different axis for the slanting, so kai would not harmonize with italic type in the same line/paragraph.

    For emphasis without emphasis dots, the “traditional” way (if I can claim this) is to pair ming with gothic (sans serif type) to emulate bold. I don’t think there’s any tradition to emulate italics.


    • Whenever I’ve needed differentiation for headings and such, I have used hei type, as I often would use sans or small caps in Western documents. I wonder, however, if hei looks well in a largely ming line: my hunch is that it looks about as tacky as modern bold forms of Garamond and Jenson (for which the bold forms ought really to be blackletter, and Baskerville really ought not to be violated with any bold forms).

      At any rate, the acquiescence to cultural imperialism displayed in the simple hitting of ‘italic’ or ‘bold’ would be astonishing to me, were I not aware of how pervasively Westernized the Chinese ethos was today. The degree to which these bad typographic practices are due to Westernization, of course, can be inflated: ignorant of decent taste, Americans use bold italic caps almost as much as Egyptians used amulets.


      • I do the same thing when I need the differentiation. That said, using gothic (when sans serif was still called gothic, meaning, “as black as blackletter”, according to what I’ve read some long time ago) for emphasis in a run of roman really is a Western invention, too. They have moved away from it somewhat, but I still find it better than artificially slanted/bolded ming. What we lack are widely available fonts with true bolds, and harmonized pairs of hei and ming fonts.

        Personally, I think the modern Chinese ethos ought to be Westernized. Unlike the Japanese, we don’t really care about our traditions.

        About hitting Bold or Italic, I think one thing some really ancient word processors did right was that they allowed us to tell them to switch to a completely different font when we hit those buttons. Unfortunately we don’t see this kind of support any more, especially when the italic of Perpetua is supposed to be called Felicity—maybe the foundries made up some names and already gotten around this oddity.


  8. »They have moved away from it somewhat, but I still find it better than artificially slanted/bolded ming.«

    By an order of magnitude.

    »What we lack are widely available fonts with true bolds, and harmonized pairs of hei and ming fonts.«

    Heartily agreed on the latter; I remain unconvinced that the former, though not a monstrosity like artificial bolding (of either ming or hei), will succeed without .

    »Personally, I think the modern Chinese ethos ought to be Westernized. Unlike the Japanese, we don’t really care about our traditions.«

    I find that some socially unhealthy patterns, including the model minority kiasu attitude, are a response to traumatic contact with the West. We also have an inferiority complex in the realm of music that drops our cultural history rather than developing and improving it organically.

    »About hitting Bold or Italic, I think one thing some really ancient word processors did right was that they allowed us to tell them to switch to a completely different font when we hit those buttons.«

    Absolutely. Centaur may have its Arrighi as italic, but my software had better not deceive me into calling forth a Centaur bold.


  9. Pingback: Fantasy | Rainbow Hill Meanders

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.